Anthropic Accuses Alibaba of Massive AI ‘Distillation Attack’ to Copy Claude Model

The American artificial intelligence company Anthropic has accused Chinese technology giant Alibaba of orchestrating a massive campaign to illicitly extract the capabilities of its advanced Claude AI models. In a letter sent to U.S. lawmakers on June 10, 2026, Anthropic detailed what it described as the "largest known distillation attack" against its systems to date.

According to the letter, operators affiliated with Alibaba and its AI research division, Qwen, used nearly 25,000 fraudulent accounts to conduct more than 28.8 million exchanges with Claude models between April 22 and June 5 of this year. The campaign allegedly targeted some of Claude's most valuable and complex functions, including software engineering and agentic reasoning capabilities.

The Technique of 'Model Distillation'

The method at the center of the allegations is known as "distillation" or "model extraction." This technique involves using the outputs of a sophisticated, pre-existing AI model to train a new, often smaller, one. By systematically querying a powerful model and feeding its responses into their own, developers can replicate some of the original model's abilities without undertaking the enormous research, development, and computational costs required to build it from scratch.

Anthropic stated in its letter that such attacks are carried out "illicitly, systematically, and at industrial scale to harvest U.S. AI capabilities across frontier labs and repackage them as their own." While the technique can be used for legitimate purposes, major AI labs' terms of service typically prohibit using their models' outputs to train competing systems.

A Pattern of Behavior and National Security Concerns

This is not the first time Anthropic has reported such activity. In February 2026, the company said it had identified similar, though smaller-scale, distillation campaigns from other Chinese AI firms, including DeepSeek, Moonshot AI, and MiniMax. The alleged operation by Alibaba is reported to be larger than all of those previous campaigns combined.

The accusations arise amid heightened U.S.-China technology tensions and growing concerns in Washington over national security. In April 2026, the White House issued a memo pledging to help American AI companies combat industrial-scale distillation efforts. Anthropic noted that the alleged Alibaba campaign occurred after this warning was issued. In response to these industry-wide concerns, a bipartisan group of U.S. senators is reportedly preparing legislation that would sanction foreign firms found to be improperly accessing and using outputs from U.S. AI models.

The issue also intersects with other U.S. government actions. Earlier in June, the Pentagon added Alibaba to a list of companies with alleged ties to the Chinese military, a designation that Alibaba denies and is currently contesting in court. As of June 25, 2026, Alibaba has not publicly responded to Anthropic's specific allegations regarding the distillation campaign.

Sources

• Inside Anthropic's Claims of Distillation Attack by Alibaba
• Anthropic accuses Alibaba of using 25,000 fake accounts to scrape Claude AI
• Anthropic accuses Alibaba of illicitly accessing Claude
• Anthropic Accuses Alibaba of Orchestrating Massive Illicit AI Cyberattack to Steal Core Capabilities
• Anthropic accuses Alibaba of 'illicitly' accessing AI models
• Anthropic notifies White House that Alibaba attempted to 'harvest' US AI capabilities: report
• Anthropic accuses Alibaba of illicitly accessing Claude AI model: Report
• Anthropic Accuses Alibaba's Qwen of Largest Claude Distillation
• Anthropic, fighting lawsuits over alleged copying of song lyrics, accuses Alibaba of copying Claude to train a rival AI

Educational and informational content. Not financial, investment, tax or legal advice, nor a recommendation to buy or sell any asset.